Artefactum 10


Conceição Casanova – Ana Lemos – Rita Araújo – Isamara Carvalho


Tracing the Historic Meaning of Two Books of Hours: The Interdisciplinary Research Underlying Conservation Decisions


The new focus on material culture draws attention to the way people and objects’ histories interact and inform each other, originating singular life stories. This article focuses on the relevance of “the biography of objects” for the conservation decision-making process, through an analysis of two Books of Hours. These important items in our cultural heritage show impressive “biographies”, as they passed on from the first owner onto their family descendants, and finally the institutional environments of this day. A team of researchers, spanning from art history to conservation science, has been revealing features of the current physical and conservation condition of two Books of Hours kept in the library of Mafra National Palace, Portugal. This work enabled the team to track down the trajectory of these objects through time, to discuss aspects of their meaning, and to put forward a biographical interpretation for each. Bearing close relation to individual “objects’ biography”, the conservation decisions and the treatment of text-block and bookbinding will be presented in this paper. Two distinct conservation solutions for each case will also be examined and compared, discussing the significance of their interpretation versus the conservation aims.



The central issue of “object’s biography” is the way that social interactions involving objects and people create meaning.[1] Books of Hours are relevant objects in our cultural heritage and can reveal very meaningful “biographies”. They were commissioned by powerful owners, passed on to family descendants, and more recently integrated in institutional environments, such as the libraries and archives of today. This trajectory has resulted in visible transformations not only in the objects’ appearance and content, but also as regards their function and value, due to changes in the social context.

In this article, the issue of the “object’s biography” is harnessed by a case study of two Books of Hours: Cofre, nrs 24 and 31. These books belong to the library of Mafra National Palace (PNM), in Portugal, and are part of a set of ten books (eight of which being French in origin, one Italian, and one Flemish). They were incorporated into the Mafra library collections in the second half of the 18th century. However, no information about their provenance was found. So important questions can be raised, such as: who brought them into this library and why? Were they purchased? Were they a gift? From whom? Such questions remain unanswered, justifying further studies about this collection.

Interdisciplinary research, involving art history and material and conservation studies, allowed us to trace the history of these objects. It also enabled us to understand the specific meaning embodied by objects throughout their existence, as well as changes in value and “significance”[2] to particular communities and societies. The study was carried out by an interdisciplinary team comprising: an art historian, mainly responsible for textual and iconographic studies; a conservation scientist, mainly responsible for the material analysis of the manuscript; a specialist in conservation and restoration of graphic documents, mainly responsible for the conservation decision-making; and two conservation students, whose task was the conservation intervention. This research was done within the framework of several academic projects.[3]

Our main aims were: (i) as already said, to trace the objects’ history; (ii) to comprehend changes in value as well as meaning, resulting from the close relationship between objects and their owners, which inform each other; (iii) to characterize the material composition and its main conservation problems; (iv) and to recover the objects’ physical integrity via a correct conservation-restoration treatment, based on the conservation decision-making process. How these objectives were reached will be explained along this article.


The context and meaning of Books of Hours

Books of Hours are private prayer books, specifically designed for lay people. Their name comes from their main section, the Hours of the Virgin, containing prayers to be said at the canonical hoursin honour of the Virgin Mary. They varied in content according to their patrons’ desires, but usually contain a calendar, the Hours of the Virgin, a standard series of readings from the Gospels, the Office for the Dead, the Penitential Psalms, several prayers and hymns.

Growing demand for Books of Hours, for private use, made this style of books of prayer hugely popular. These books first began to appear in the 13th century and by the later Middle Ages were widespread. They were produced in the workshops of large European urban centres, such as Paris, from the 13th century until the early 16th century, contributing to the development of a class of professionals, among whom booksellers and illuminators.[4] The studios/workshops of the urban centres, namely in France and Flanders, were led by a master who was usually responsible for the most important illuminations in the book, plus a group of artists under his guidance.[5] In the early 15th century, Books of Hours become an indispensable item among the wealthy.[6] The decoration and miniatures are important components. Typically, they represent the natural world and scenes of daily life and religion and performed quite a practical function, marking sections out and articulating the different parts of the manuscript. The miniatures were commonly used in the opening of the offices, acting as a visual call and a stimulus to meditation.[7]

In the case of commissioned books, images also served to invite the owner into the sacred space: he or she were represented in a gesture of prayer and kneeling, next to the sacred figures, such as the Virgin and Child. That way they reaffirmed their faith, emphasizing the religious experience as a unique and private moment. This kind of representation also exhibits one of the main characteristics of Books of Hours, as they were produced for a particular individual.[8] Books of Hours were small in size, so as to be carried easily, but they could be quite luxurious as well. As manuscripts, gatherings – composed of folios of thin parchment – form the text-block (Figure 1a), which could be extensively illuminated, using a rich palette of pigments. The original bindings in “Gothic style” include a sewing structure on double raised thongs with headbands, laced into the wood boards, covered in precious textiles or decorated tanned leathers, and fastened with metal clasps. The spine could be slightly rounded and reinforced with parchment strips between the panels (Figure 1b).


FIGURE 1a. Text-block construction.

FIGURE 1b. “Gothic style” bookbinding construction.

The selection of texts and the iconographic program, the richness of the ornamental decoration and miniatures, and the quality of materials used vary widely, according to the economic means and personal taste of the commissioners. These different features also represent the commissioners’ spiritual needs and/or their social standing.[9]


Presentation of the case studies

These two Books of Hours from the 15th century are of French origin, most probably from Autun[10] and Paris.[11] They probably belonged to the wealthy, given the select materials and rich colour palette used, including gold, silver, lapis lazuli and other expensive pigments.[12] They were also altered along their existence, something that was quite common: Books of Hours were usually kept in the same family for several generations and, according to the new owner’s perspective and taste, changes could be made – including on the text-block, where new prayers and miniatures might be added or removed, sometimes requiring the book to be bound afresh. The latter could be necessary because of text-block alterations, heavy use (conservation reasons), or simply out of the need to enrich it or to make it more consistent with the new environment. One or several of these reasons could lead to structural changes in the binding. For this reason, many of the early books that have survived to this day were heavily altered or show much later bindings.[13] This is the case with these two Books of Hours, which became part of the PNM and were given completely new bindings.[14] These bindings were not compatible with the 15th-century text-blocks, and both books were in a very poor conservation condition needing urgent intervention.


Text-block analysis

The original text-block of manuscript Cofre nr 24 dates from around 1420, with additions dating from around 1470.[15] The text-block is inscribed on good quality parchment whose margins were trimmed probably during re-binding, affecting slightly both the text and the decoration on the margins. The manuscript is formed of 23 gatherings totalling 181 folios generally with 14 lines of text each, written in Latin and French. There are fourteen full-page miniatures with three, four or five lines underneath. It contains a calendar in French and offices and prayers, mostly in Latin, but some in French. As yet, we do not have information to identify all of the saints included in the calendar of Cofre nr 24. Saint Benigne, featured on the 24th of November, is one of the first evangelizers in Burgundy, namely in Autun. So, as already mentioned, Cofre nr 24 is a Book of Hours with the Office of the Virgin that probably comes from Autun. Although the text uses the male formulation “famulo tuo” instead of the female formulation “famue tue”, two miniatures confirm that this Book of Hours belonged to a woman: in the one opening the prayer “Obsecro te” (Figure 2a, fl.17v) the owner dressed in woman’s clothes is represented kneeling in prayer before the Virgin and Child, accompanied by two other female figures (most likely the daughters). Also, between gatherings 19-20, in a folio added later (probably in the second half of the fifteenth century), that illustrates a prayer to Saint Anne, “De sancta anna antiphona” (Figure 2b, fl.153r), the owner is equally represented kneeling, holding an open book in front of St. Anne who is teaching the Virgin how to read; thus witnessing the transmission of this book between owners and from generation to generation.[16]


FIGURE 2a. PNM, Cofre nr 24, fl.17v, Virgin with Child and donor, shown probably with her daughters, miniature above 4 lines of text, 92x64 mm.
FIGURE 2b. PNM, Cofre nr 24, fl.153r, Saint Anne teaching the Virgin how to read and donor, miniature over 4 lines of text, 97x71 mm.

Cofre nr 24 also presents other peculiarities. In the set of texts there were material losses, namely at the end of the Matins of the Hours of the Virgin the text on fl.39 is unfinished and the added folio which begins at Lauds does not fill the reported fault. Indeed, the miniature of the Visitation (Figure 3a, fl.39v), which is quite different from the rest, belongs to a later period (c. 1470), since the oldest part of the manuscript dates from about 1420. We found the same at the beginning of the Office of the “Sancto spiritu”, in which the miniature trace of Pentecost (Figure 3b, fl.112r) suggests the same artist hand in the previous one. In this case, the folios which contained the miniatures of the Visitation and Pentecost were lost, the owner deciding to compensate this by having the miniatures that are currently found with the missing text part executed or acquired. The insertion of the miniature onto a folio composed of a collage type margin is also worth mentioning. We know only two other similar cases: a miniature representing the birth of the Virgin found in a drawings cabinet at the Louvre[17] and a miniature representing the Virgin born and other figures from a manuscript found at the council library of Amiens.[18]


FIGURE 3a. PNM, Cofre nr 24, fl.39v, Visitation miniature above 3 lines of text, 86x50 mm.
FIGURE 3b. PNM, Cofre nr 24, fl.112r, Pentecost miniature above 3 lines of text, 86x50 mm.

However, these are not the only flaws. Indeed, the complete study of the texts of Cofre nr 24 allowed us to identify other missing parts: in the Gospel Lessons of Luke, Matthew and Mark the beginning of the texts and the miniatures are missing; and in the Hours of the Cross the beginning of the text is lacking (there is no continuity between folio 108v and folio 109r).[19] The manuscript’s original text-block in Cofre nr 31 dates from around 1440, with additions from around 1490.[20] The text-block on parchment was also trimmed on the margins probably during re-binding, slightly affecting the text and the decoration in the margins. The manuscript is composed of a calendar, in French, in the use of Paris,[21] of offices and prayers (mostly in Latin, but also in French);[22] the Hours of the Virgin and the Office of the Dead also accordig to the use of Paris. This confirms the region as the origin of this manuscript. The manuscript is organized in 23 gatherings totalling 162 folios and it contains three full page miniatures with four lines underneath, seven small miniatures, many illuminated initials (major and minor), line fillers and other decoration in the borders with golden ivy leaves, stylized flowers and acanthus.[23]

In Cofre nr 31 we identified three different artists, corresponding to different techniques of colour application and handwriting. The stylistic analysis of the miniatures led to the identification of three groups, namely: (i) one formed of the three full-page miniatures over a four-line text, the Annunciation (Figure 4a, fl.25r), David in Prayer (fl.73r), and the Pentecost (fl.91r), which displays a similar frame ending with a segmental arch and identical margins, showing an identical decorative pattern; (ii) another formed of the historiated initial of St Cristopher (Figure 4b, fl.141v) and the miniature of St Catherine (fl.147v), with an identical decorative pattern, measuring 5 and 6 spaces height respectively, inserted in the text of the Suffrages of the Saints; (iii) and a third group that corresponds to a later addition, composed of historiated illuminations inserted in pictures: again St Catherine (Figure 4c, fl.149r), confirming the devotion of the owner of the manuscript to this saint, the Virgin and Child (fl.153r), the Pietà (fl.154r), the Virgin surrounded by four angels (fl.155r), and the Virgin in Prayer (fl.160r). This third decoration typology also contains an ornamented border on the right along the text height, and the picture inserted in a pink/purple frame with a gold fillet.

FIGURE 4a. PNM, Cofre nr 31, fl.25r, Annunciation miniature above 4 lines of text, 85x65 mm.
FIGURE 4b. PNM, Cofre nr 31, fl.141v, historiated initial of S. Cristopher.
FIGURE 4c. PNM, Cofre nr 31, fl.149r, miniature of St Catherine.

In both books the colour composition, namely the pigments applied, was studied through applied methods, such as: optical microscopy, micro energy dispersive X-ray fluorescence (µ-EDXRF), fibre optics reflectance spectroscopy (FORS), and micro Raman spectroscopy (µ-Raman); and through the micro-invasive technique of micro Fourier Transform Infrared spectroscopy (µ-FTIR) for the paint formulations characterization. The results of the scientific analysis with the different techniques shows a molecular palette in accordance with that expected for the period. We found lapis lazuli, azurite and indigo for blues, malachite and basic copper sulphate for greens, mosaic gold and lead-tin yellow for yellows, vermillion, minium and brazil wood lake for reds and pinks, lead white for light tones and carbon black for dark tones and lines, as well as burnished gold and silver. Egg white and Arabic gum were used as binders.[24] The analyses using these different techniques also allow us to confirm differences in the inks and the pigment palette on the three groups of illuminations identified for Cofre nr 31, corroborating the study of the iconography. The iron gall writing inks analysed through μ-EDXRF have three different compositions that correspond to three groups of folios, thus confirming three separate scripts.[25] It was also observed that the miniatures of the first group were composed of lapis lazuli over azurite and indigo to create the shades of the robes, but for the figures of the second and third groups only azurite was applied. The difference increased when analysing oranges and reds: in the first group, the mantle of King David (fl.73r) is painted with minium and an organic lake, but the strawberries, flowers and acanthus leaves of the borders are a mixture of minium and vermilion with a carmine glaze of an organic lake; while for the miniatures in the second and third groups, the artists differ from the first group by using vermilion only for all the reds. This approach allowed us to identify two techniques of pigment application, distinguishing the artist of the first group of illuminations from the other two.[26]


Material structure and bookbinding

Besides the additions and subtractions found in the text-block of the two books previously mentioned, another material transformation concerns the replacement of the bookbinding. In fact, the books were rebound according to the local library style sometime between the late 18th and the early 19th centuries, reflecting the meaning and function changes of books over time. Custodian institutions often changed the bindings according to their policies and rules. This was the case with these two books, which displayed decorative features, techniques and materials alike to the entire collection, including the ten Books of Hours in the PNM Library.[27] The new bindings brought books in line with the setting of the Library, giving a harmonious aesthetic appearance to the shelves that cover the walls of this majestic 18th-century library, ordered by the absolute ruler João V of Portugal, to convey a message of power and wealth of the kingdom. This context justifies the presence of a bookbinding workshop at PNM, where we believe the books where rebound.[28]

The books reveal two-up sewing on raised cords (Cofre nr 24 on five, Cofre nr 31 on four) and simple endbands on two colours. The cords were attached to flax pasteboards by lacing through a single hole. Pasteboards where covered in full calf leather, decorated with sprinkling, a technique that was fashionable in the period. However, the two books show different application techniques: one was done by means of a brush knocked against an iron bar (Cofre nr 24); and the other applied by using a sponge (Cofre nr 31). This probably reveals different bookbinders’ hands. Nevertheless, in both cases the skin turn-in includes cutting along the endbands and shows skin overlapping at the corner, which was probably the work of a same workshop. The endpapers are formed of a four-leaf unit of rag paper, where one of the leaves was coarsely torn forming an endleaf-hook. The latter plus a complete leaf were used as pastedowns and the other two complete leaves were used as flyleaves.

The spine was reinforced over the endbands with a parchment strip and between panels with recycled paper strip, which reveals printed Portuguese characters.[29] It was also decorated with gold tooling to call the viewer’s attention to the shelves. All the ten books show an inscription on the second panel of the spine (“Heures Devotes” on Cofre nr 24 and “Officivm Parv. B.V.” in Latin for Virgin Hours on Cofre nr 31). Finally – apart from Cofre nr 24 due to a missing area on the leather of the spine cover – the books have the lettering “MAFRA” in gold tooling on the bottom of the spine, again suggesting they were all rebound at the workshop of the PNM Library.[30]


Conservation condition study

The two Books of Hours were in a very poor state of conservation. Concerning the condition of the bookbinding, the main damage consisted of heavy deformation and the breakdown of the sewing structure. This also affected the text-block conservation condition. Parchment is a very hygroscopic support material, changing its dimension and shape according to oscillations in relative humidity. Humidity and pollutants can easily penetrate a book with the broken sewing structure and no other system to hold it in place. Therefore cockling and distortion of the parchment support occurred, as well as alterations in the medium and the pictorial layers. Pigments flaking and several oxidation processes were reported, such as silver leaf oxidation, mainly applied to the miniatures, are discussed below in this article.[31]


Text-block condition

The writing inks used in the text of light to dark brown colour tones, are mainly composed of iron, as well as copper and zinc, as indicated by μ-EDXRF analyses of Cofre nr 31. Although the presence of three different inks can be recognized (with these elements found in different proportions), generally speaking it corresponds to quite stable gall inks. Very few areas of ink fading are found.[32] Regarding the miniatures, the condition is generally acceptable but certain pigments are brittle and loose. This is the case with the green found in Cofre nr 24, composed by malachite and a basic copper sulphate (possibly brochantite), the blue composed of lapis lazuli and azurite, and the lead white (Figure 5a).[33] This degradation may be partly related to the bookbinding condition and the parchment deformation, but may also be connected to the material composition of colours and the applying techniques. Several authors have mentioned the grain size of certain pigments and/or the low amount of binder.[34] However, to be conclusive on this matter further investigation is necessary.

The detachment of painting material can also relate to mechanical factors arising from the heavy use of this type of manuscript. Other signs of intense use are connected to the partial dissolution, spreading and dragging of paints, due to water-solubility when in contact with moisture. This is the case of certain borders and some historiated initials that are blurred.[35] These deterioration processes are usually an indication of intense usage of specific folios of the manuscripts.

The gold leaf also shows some surface wear, but the strongest deterioration process concerns the silver leaf oxidation.[36] Its oxidation led to a dramatic aesthetic transformation, since the glossy, smooth metallic surface has become a rough dark colour surface, changing the original appearance of the image and its expressive qualities (Figure 5b). Another rather unpleasant effect of this deterioration process is the migration of pigments to the other side of the folio. The phenomenon can be very forceful with silver but varies greatly depending on the pigment colour. At Cofre nr 31 some reds mixing vermilion and minium also show similar degradation, with the reds generally darkening into brown and colour migration to the verso probably due to the mixture of mercury sulphide, as reported by Carvalho.[37] The migration phenomenon is also observed in green and blue areas, namely a greenish tone that is probably associated to malachite and azurite pigments. Also, the acanthus leaves and flowers coloured with powdered gold showed an orange-yellow tone on the back of the folio (Figure 5c).

FIGURE 5a. PNM, Cofre nr 31, detail of loose pigment, fl.25r (25x).
FIGURE 5b. PNM, Cofre nr 31, detail of silver oxidation, fl.91r (32x).
FIGURE 5c. PNM, Cofre nr 31, detail of pigments migration fl.91v (7x).

Regarding the support assessment, the parchment conserves it full integrity. However, backfolds of the outer folios of book gatherings showed strong damage caused by the animal glue used on the strengthening and shaping of the spine. This is particularly evident in Cofre nr 31, where every fold exhibits parchment lost and oxidation stains. The endpapers were also heavily glued. Nonetheless, the assessment of Cofre nr 31 using the IDAP protocol (Improved Damage Assessment of Parchment) reveals parchment deformation, superficial dirt, discolouration and illumination damage as the main conservation problems.[38]


Bookbinding condition

In both books the binding was deformed and dismantled, showing loose folios and gatherings. A rather tight binding (characteristic of the period of manufacture), with a round and heavily glued spine, plus high tension in the sewing line, caused holes and tears on the backfold of the folios of the gatherings. Also, it caused the sewing to break, the deformation of cords and endbands, and consequently the tearing and loosening of the spine, especially along the joints. The skin on the spine and over the pasteboards also showed intense surface wear and some material loss and insect holes, among other signs of degradation and less visible deterioration features, such as skin cracking and dryness.

Apparently, these bindings between the late 18th and the early 19th centuries were not fulfilling their protection function (Figure 6a, Figure 6b), contributing to the poor conservation condition of the text-block. We must realise, however, that rebinding was done probably because the previous binding were already in a bad shape. The heavy handling of Books of Hours and their frequent alteration justifies the poor condition of bindings.

FIGURE 6a. PNM, Cofre nr 24 before treatment: front and back covers, bottom edge, spine and general view.
FIGURE 6b. PNM, Cofre nr 31: front and back covers, bottom edge, front edge and general view.

Conservation process

Due to the conservation condition as previously described, it was essential to treat all the structure to conserve these two manuscripts. The intervention focused on the recovery of the bindings, maintaining their integrity and preserving all traces of the history of each individual book. Nonetheless, the conservation needs of the whole set were taken into consideration to guarantee long-term preservation conditions. In what concerns the repairing process of text-blocks the principle of minimal intervention was followed.


Text-block conservation

In both cases it was necessary to dismantle the quire, dry clean all the set with a soft brush and the backfolds with a smoke sponge®. Special attention was given to the backfolds of folios were fungi were detected. In these cases the area was vacuum cleaned with a HEPA filter®. Small traces of adhesive were also dry cleaned with the help of a hard point instrument whenever necessary. Also, when the adhesive was hardened and dry, it was smoothed with a slightly moistened cotton swab in a solution of 50% deionized water and 50% ethanol. After the cleaning process, creases and cockling of the folios were removed or attenuated using gradual and controlled humidification. The creases caused by the backing and rounding of the spine, done during the 18th or 19th-century binding process, were also attenuated during this process. Finally very slight pressure was applied between heavy felts, blot-papers and wooden boards.

The following step consisted in repairing a few tears and filling in the existing gaps and material losses, mainly caused by the previous tight sewing and handling. This was necessary before the re-sewing process of the gatherings on raised cords. The damaged outer backfolds of folios of the gatherings were locally reinforced with thin parchment and the inner ones with synthetic collagen, using purified wheat starch paste. The choice of adhesive was based on the conservators’ experience, taking into consideration the adhesive qualities: strong enough for the type of materials involved, smooth and flexible enough to apply in even layers, relatively slow drying, as well as having reasonable physical and chemical stability, exhibiting no significant colour alteration, and showing good reversibility properties with ageing, in the long run.[39]

In order to recover the balance of the whole text-block it was considered replacing some of the single missing leafs, which were removed over time. Whenever an entire folio was missing nothing was done, respecting and understanding the changes that the item incorporated along its existence, as part of its own identity. Missing pages were replaced by a new blank parchment folio in a clear and visible manner, respecting the authenticity of the item and its individual history. That way each new insertion is clearly detectable, showing the relevance of the object’s biography for the conservation decision and treatment methodologies. A new sheet of parchment, slightly thinner than the original folios but with similar characteristics, was used for the operation. The new parchment was analysed by FTIR ATR to make sure that no harmful elements or products were used in the treatment of the modern parchment leaf, but it must be said that it proved very difficult to find a suitable parchment in the conservation products market.

It is important to point out that no intervention was done regarding the illumination, avoiding any alterations to the text-block original materiality and its 15th-century identity. We believe that improving the book structure and binding and further preventive conservation will be enough to recover the stability of the text-block and miniatures, avoiding the use of products for consolidation or treatments that are more invasive.[40] The item’s conservation condition will be monitored to check if the preservation conditions are sufficient to avoid further damage.


Bookbinding conservation: two different approaches

The books were dismantled, so a complete re-binding process was necessary. At this stage, ethical concerns are always present especially when fairly recent bindings do not fulfil the conservation needs of an antique text-block completely.[41] The process requires a systematic and full examination, layer by layer, to signal, characterize and register all the material evidence that will usually reveal the overlapping of layers from different historical moments. We start by looking for signs of different sewing holes in the middle of gatherings of the text-block and the spine, the analysis of the existing sewing structure and its supports, endband typology when present, spine features such as the characteristics and shape of linings, the boards’ material and shape, the lacing patterns and sewing supporting the attachment of boards, the cover material and covering method, the characteristics of endleafs, and finally any remaining decorative features.

Two main options can be raised: (i) to recover the surviving materials and techniques as much as possible, which requires reaching a balance between conservation needs and historical veracity; (ii) more radically, to create a new binding with better preservation qualities.[42] The last option should be very well substantiated and discussed by an interdisciplinary team. The last word should be the book’s owner’s or trustee’s. Previous discussion is essential because this option often results in a pastiche, since conservators tend to use technical solutions contemporary to the text-block. Apart from the final conservation decision, the following main steps have to be considered: to register, clean and conserve the bookbinding materials found and decide which can be re-used, substituted and/or kept for future memory; to foliate the text-block in position by collecting gatherings and deciding on the re-sewing methodology and the endbands; considering spine linings and shaping; to decide on the re-usage of boards and the attachment technique; to decide on the covering material re-used and the technique applied and finally to decide on the endleafs re-usage and their attachment methodology.

In general, the decision was made to re-use materials and maintain the 18th or 19th-century bindings found in these books, preserving the appearance of the collection of the ten Books of Hours in the PNM. Following the 18th and 19th-century binding fashion, a new sewing on raised cords and new endbands of two-colour silk thread over cord were made, following the traces found in place. Pasteboards and cover leather were re-used, but a new reinforcing leather spine was applied, under the 18th or 19th-century full cover. The board attachments were also reproduced in the same fashion as the previous binding: the cords were laced through a simple hole and stuck to the boards. After the conservation treatment of the four-leaf endleaf unit, it was also stuck in place in the same fashion as the previous binding found: an endleaf-hook plus a complete leaf were used as pastedowns; and the other two complete leaves used as fly leaves.

In terms of the structure, namely the sewing methodology applied on the gatherings of the text-block and spine reinforcement, a different conservation approach was taken for the two case studies (Figure 7a and 7b):

a) In Cofre nr 24, the state of conservation and reasonable consistence of backfolds allowed us to follow the sewing structure of the 18th or 19th-century binding, namely the two-up sewing with bypass and the spine linings with handmade paper at panels, between the raised cords;

b) In Cofre nr 31, taking into consideration the heavy damage of the backfold of the outer folio of the different gatherings and the weakness of the inner fold of the folio in almost all of the gatherings, we decided to change the sewing structure of the previous binding for a traditionally medieval sewing: sewing all along in every section; a new system of skin attachment was also implemented: the new spine leather used for reinforcement was attached to the cords only with few stitches and no adhesive was used.[43]

FIGURE 7a. PNM, Cofre nr 24 spine reinforcement.
FIGURE 7b. PNM, Cofre nr 31 spine stitch attachment.

With Cofre nr 24 we came up with three layers of adhesive: the first one comprised paper linings between spine panels, the second one a new reinforcing leather spine; and the third one the old skin leather, stuck on the top of the new one. With Cofre nr 31 we reduced this to one, due to the frailty of the backfolds of gatherings. Again, a wheat starch paste was used only were the 18th or 19th-century leather of the spine was stuck over the new reinforcing leather spine. This was an important technical improvement in terms of conservation, easing the opening of the book and avoiding the hardening of the adhesive that usually results in opening difficulties and leads to spine deterioration over the years.

Again due to different conservation conditions, the conservation methodology used for the treatment of the paper of the endleaves was slightly different: in Cofre nr 24 an ultrasonic humidifier with a solution of deionized water and ethanol (1:1) was applied[44] and was sufficient to lift the paper with the help of a scalper; in Cofre nr 31 the delicateness of paper and heavy attachment of the pastedowns to the boards required the use of a gel composed of carboxymethyl cellulose (CMC). In both cases the endleaves were washed in a bath with deionized water and whenever necessary deacidified with a calcium carbonate solution, leaving an alkaline reserve and a pH of around 8 (Figure 8a and 8b).

FIGURE 8a. PNM, Cofre nr 31 endleaves first bath.
FIGURE 8b. PNM, Cofre nr 31 endleaves second bath.

Final discussion and conclusions

In aesthetic terms we can say that no obvious change is visible in the two case studies: in both cases we ended up with a conserved-restored late 18th or 19th-century type of binding (Figure 9a and 9b). Nevertheless, in conservation terms, we believe that the structural changes of the conservation process, especially as applied on Cofre nr 31, will improve longevity, allowing a safe and easier handling of the books in future.

FIGURE 9a. PNM, Cofre nr 24 after conservation treatment: front and back covers, bottom edge, spine, general view.
FIGURE 9b. PNM, Cofre nr 31 after conservation treatment: front and back covers, bottom edge, spine and general view.

In ethical terms, both items are “re-treatable”.[45] All evidence, every detail and all changes made during treatment were recorded as a matter of fact, following current recommendations. Also, this was the only way to guarantee the full interpretation and maintenance of the “cultural significance” of these objects for future generations. Current conservator's responsibilities imply an understanding of changes in the perceived value of objects, as well as of specific meanings based on their singular history. Thus consistent conservation decisions can be made. Also objects keep their ability to inspire and enrich people’s lives and experiences, as well as to reveal the past and reflect the dialogue between people and objects and the diversity of our irreplaceable cultures and identities.

Comparing conservation decisions, two different solutions were found, which correspond to different conservation conditions. The heavier damage of outer backfolds of gatherings and, consequently, of the spine of Cofre nr 31 implies a different approach in this case. This proves that, although new changes in function or meaning of these items are expected to happen in the near future, according to the deterioration path and conservation needs each object will still follow “its own way” and have a specific “object biography” – one which can be visible and detectable in material terms. In other words, new social interactions are always being created between objects and different actors and, necessarily, objects acquire new meanings and value along their existence, which in turn are reflected on the material appearance and singular history. Therefore, as we have tried to prove with these two case studies, the conservation decision-making process requires a comprehensive knowledge about objects, only possible through the work of a large team covering art history, technical art history, hard science, and conservation science, so as to fully understand the singular history of the object, its material composition and deterioration processes, and its conservation state and needs. This results in a completely “re-treatable” conservation process, based on the interpretation of historical biography.

In fact, a long time ago (in the 19th century and the first half of the 20th century) researchers in the humanities, historians in particular, used to impose their beliefs and decisions on conservators. Since the second half of the 20th century, the importance of hard sciences for the conservation work has gained recognition – these case studies proving the advantages of having a conservation scientist in the team to understand deterioration processes and relate them with the material composition and its behaviour. It is now time to recognise the importance of social science for conservation, allowing for the full interpretation of objects through their biographic study. This should be done previously to the conservation decision, as a way to ensure high standards of conservation.

As we have seen, Books of Hours are important objects of our cultural heritage, having inspiring “objects biographies”, revealing historic events, and leading us to different spaces and people. The two books have different origins, one (probably) belonging to the use of Autun and the other to the use of Paris. But when we look closely at the material evidence in connection with immaterial interpretations much more can be revealed. In Cofre nr 24, a rich book exhibiting fourteen full-page miniatures, the dialogue existing between the commissioners and the object, in the beginning and end of 15th century, can be understood in full: probably two female owners expressed their beliefs, at different times, by making themselves represented in the book. In the case of Cofre nr 31 the dialogue with another character, the artist illuminator, is more evident. Three different artists’ hands are perceptible: two different illuminators, probably from the same workshop where the book was made; and a third one added when a new owner decided to make additions, reinforcing his faith and evidencing his personal taste. But both books ended up being kept and valued by the same institutional keeper: the Library of Mafra National Palace. The dating of bookbinding acts corresponding to the Library style, and the heavy use and damage of Cofre nr 31 when compared to Cofre nr 24 are different dialogues and need further clarification.

We hope that the dialogue, as mediated by the interdisciplinary team, between the objects and the conservator has been comprehensibly exposed and debated in this article, revealing the conservators' thoughts and decisions versus the objects' specificity and history.



Appelbaum, Barbara 1987: “Criteria for Treatment: Reversibility”. P. 65-73. Journal of the American Institute for Conservation, Vol. 26, No. 2 (Autumn).


Araújo, R., Casanova, C., Melo, M. J., Lemos, A. 2015: “Ethical and Technical Concerns during the Conservation Process of a Religious Book. The Book of Hours from the Library of Palácio Nacional de Mafra”. P. 241-252. European Journal of Science and Theology, Vol.11, No.2 (April).


Araújo, R. 2017: “Examining the books of hours in Portuguese collections”. Nataša Golob, Jedert Vodopivec (Eds.). P.205-226. Bibliologia 45. Bookbindings Theoretical Approaches and Practical Solutions. Turnhout: Brepols Publishers. Available: (01-09-2018)


Araújo, R., Nabais, P., Pombo Cardoso, I., Casanova, C., Lemos, A., Melo, M..J. 2018: “Silver paints in medieval manuscripts: a first molecular survey into their degradation”. P.1-13. Heritage Science 6 (8). DOI: 10.1186/s40494-018-0172-7.


Carvalho, I. 2015: Processos de degradação em manuscritos iluminados: estudo, conservação e restauro de um livro de horas do século XV do Palácio Nacional de Mafra. Tese de mestrado. Lisboa: Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa. Available: (25-08-2018).


Carvalho, I., Casanova, C., Araújo, R., Lemos, A. 2018: “Colour identification, degradation processes and findings in a fifteenth-century Book of Hours: the case study of Cofre nº31 from Mafra National Palace”. P. 1-17. Heritage Science,. 6 (9). DOI:10.1186/s40494-018-0174-5.


Casanova, C. 1995: “Recuperar livros antigos: um problema ético e técnico”. P.61-62. Do bisturi ao Laser. Lisbon: Fundação Calouste Gulbenkian.


Casanova, C. 2001: “Restauro de 100 Chancelarias Régias – Um caso exemplar de conservação”. P.32-36. Boletim ADCR, nº10/11. Lisbon: Associação para o Desenvolvimento da Conservação e Restauro.


Casanova, C. 2003: “Contribuições para a conservação de Forais Manuelinos”. P.177-186. Vária Escrita, Cadernos de Estudos Arquivísticos, Históricos e Documentais, nº10. Sintra: Câmara Municipal de Sintra.


De Hamel, C. 1986: A History of Illuminated Manuscripts. Oxford: Phaidon Press.


Gosden, Chris, Marshall, Yvonne 1999: “The cultural biography of objects”. P.169-178. World Archaeology, Vol. 31, No. 2 (October), The Cultural Biography of Objects.


Lawson, M., Yamazaki-Kleps, A. 2002: “Examination and conservation of the fifteenth-century parchment manuscript, the Belles Heures of Jean, Duke of Berry”. P. 129–34. Works of art on paper: books, documents and photographs: techniques and conservation, Contributions to the Baltimore Congress, 2-6 (September).


Lemos, A. 2012: Os livros de horas iluminados do Palácio Nacional de Mafra. Lisboa: Instituto de Estudos Medievais/Faculdade de Ciências Sociais e Humanas - Universidade Nova de Lisboa and Palácio Nacional de Mafra.


Lemos, A., Araújo, R., Casanova, C. 2015: “O cofre nº 24. Um livro de horas do Palácio Nacional de Mafra: caso de estudo e de intervenção”. P. 82-93. INVENIRE special issue.


Musée du Louvre 2011: Les Enluminures du Louvre. Moyen âge e Renaissance. Paris: Éditions Hazan.


Marrow, JH.1995: As Horas de Margarida de Cleves. Lisbon: Museu Calouste Gulbenkian.


Melo, M. J., Araújo, R., Muralha, V. S. F. 2012: “O que nos dizem os materiais da cor sobre os livros de horas do Palácio Nacional de Mafra?”. P.103-115. Os Livros de Horas Iluminados do Palácio Nacional de Mafra, ed. by Ana Lemos, Lisbon: Instituto de Estudos Medievais /Faculdade de Ciências Sociais e Humanas - Universidade Nova de Lisboa and Palácio Nacional de Mafra.


Melo, M. J., Otero, V., Vitorino, T., Araújo, R., Muralha, V., Lemos A., Picollo, M. 2014: “A Spectroscopic Study of Brazilwood Paints in Medieval Books of Hours”. P. 434-444. Applied Spectroscopy, 68 (4). DOI:10.1366/13-07253.


Melo, M. J., Araújo, R., Castro, R., Casanova, C. 2016: “Colour degradation in medieval manuscripts”. P. 837-844. Microchemical Journal 124.

DOI: 10.1016/j.microc.2015.10.014


Ottosen, Knud. 2007. The responsories and versicles of the latin office of the dead. Copenhagen.


Phelan, W. H., Baer, N. S. and Indictor, N. 1971: “An Evaluation of Adhesives for Use in Paper Conservation.” P. 58-75. Bulletin of the American Group. International Institute for Conservation of Historic and Artistic Works, Vol. 11, No. 2 (April).


The Burra Charter. 2013: The Australia ICOMOS Charter for Places of Cultural Significance. Available: (01-09-2018).


Van Asperen de Boer, J. R. J. 1974: “An Examination of Particle Size Distributions of Azurite and Natural Ultramarine in Some Early Netherlandish Paintings.” P.233-43. Studies in Conservation, Vol.19, Nº 4 (November).


Wieck, R. S. 1988: Time Sanctified: The Book of Hours in Medieval Art and Life. London: Sotheby’s Publications.


[1] Gosden & Marshall 1999: 169-70.

[2] In the sense of the “cultural significance” concept in The Burra Charter by the Australia ICOMOS charter for places (2013), referring to the capacity of heritage to “enrich people’s lives, often providing a deep and inspiring sense of connection to community (...), to the past and to live experiences”. (accessible at: - 01-09-2018).

[3] This work has been mainly developed under the following framework: “The Books of Hours in the Portuguese collections: matter, form and meaning”, the PhD project of Rita Araújo, supervised by Casanova and co-supervised by Melo and Lemos, which follows from her MA studies; the “Catalogue of French illuminated Book of Hours in the Portuguese public collections” the PhD project of Lemos, supervised by Miranda and co-supervised by Stirnemann and Melo; plus “O Cofre nº 31 do Palácio Nacional de Mafra: estudo, conservação e restauro de um livro de horas do século XV”, the MA dissertation by Carvalho, supervised by Casanova and co-supervised by Araújo and Lemos. The technical description made during the conservation treatments of both Books of Hours and some of the results about the binding typology characterization were first presented in the following paper: “ ‘Coffer nº24’ and ‘coffer nº31’ conservation, two books of hours conserved in the Nacional Library of Mafra”, by Lemos and Araújo, read as part of the panel “The Conservation-Restoration and the Art History: decision making and preservation methodologies for cultural heritage”, organized by Casanova, at the 4th International Conference – Medieval Europe In Motion - The Middle Ages: a Global Context?, in Lisbon, 13-15 of December, 2017.

[4] De Hamel 1986: 168-69.

[5] Wieck 1988: 28-30.

[6] Wieck 1988: 27.

[7] Marrow 1995: 17.

[8] Lemos 2012: 27.

[9] Wieck 1988: 33.

[10] Lemos 2012: 63-65, 125; Lemos et al. 2015: 85, 87, 88. The study of the identified sequences of the Office of the Virgin use point to the region of Autun, with only four variants being noted; regarding the Office of the Dead and taking into account Knud Ottosen’s work, it was not possible to identify the usage of this Book of Hours for sure.

[11] Lemos 2012: 125-26, 130.

[12] Araújo et al. 2018: 3-4; Carvalho et al. 2018: 10; Melo et al. 2014: 435, 440-41.

[13] Casanova 2001: 35-36; Casanova 2003: 180.

[14] Araújo et al. 2015: 131; Carvalho et al. 2018: 2; Araújo 2017: 210.

[15] Lemos 2012: 125-26.

[16] Lemos 2012: 99; Lemos et al. 2015: 84. So far, no explanation has been found for the shaven face of the figure at the right of the composition.

[17] Les Enluminures, 2011: 206-207. Fragment from a Gradual conserved in the cabinet of drawings of the Museum of the Louvre, reference: RF 29080.

[18] Manuscript reference: BM – ms. 0107 – the authors thank Joris C. Heyder (Freie Universität Berlin) for information about this manuscript.

[19] Lemos, 2012: 125; Lemos et al. 2015: 85-6, 88.

[20] Lemos 2012: 130.

[21]  Lemos 2012: 33-5, 137-51. Note the presence of Saint Geneviève, patron of Paris on the 3rd of January and the 26th of November. For more information see Lemos, Os livros de horas iluminados do Palácio Nacional de Mafra.

[22] Lemos 2012: 130. The beginning of the text is missing in the following prayers: Fifteen Joys of the Virgin (fls. 136-138v) and the series of Seven Requests to Our Lord (fls. 139-140v). These two, plus the prayer to the Holy Sacrament (fls. 161-162v), are in French. The prayer of “Les Sept vers Saint Bernard” (fls. 151v-152v) has the rubric in French and the text in Latin.

[23] Lemos 2012: 130.

[24] Araújo et al. 2018: 7-10; Carvalho et al. 2018: 3-10.

[25] Carvalho et al. 2018: 4.

[26] Carvalho et al., 2018: 13.

[27] Most of the PNM bindings exhibit similar aesthetic and technical features as first noticed by Casanova and presented with Lemos (historical context) and Araújo (colour formulation) as the paper “ Regards croisés des historiens de l’art, les conservateurs-restaurateurs et les chimistes: une étude de cas”, in the homage session for Michel Pastoureau at the Institut Français du Portugal, Lisbon, 19th of September 2014, organized by Lemos and sponsored by the Instituto de Estudos Medievais, the Instituto Português de Heráldica, the Institut Français du Portugal and the French School Charles Lepierre in Lisbon.

[28] Araújo et al. 2015: 131; Araújo 2017: 210.

[29] Araújo et al. 2015: 131.

[30] Araújo et al. 2015:131.

[31]  Melo et al. 2016: 840; Araújo et al. 2015: 133-35.

[32] Carvalho et al. 2018: 4; Araújo et al. 2015: 133.

[33] Carvalho et al. 2018: 11-12; Araújo et al. 2015:134-35.

[34] Van Asperen de Boer 1974: 233-34; Lawson et al. 2002: 131-32; Melo et al. 2012: 112-13; Melo et al. 2016: 840-41; Carvalho et al. 2018: 11-12.

[35] Carvalho et al. 2018:12.

[36] Araújo et al. 2018: 1; Carvalho et al. 2018: 10.

[37] Carvalho 2018: 10-11.

[38] Carvalho 2015: 15; Carvalho et al. 2018: 2, 4.

[39] Phelan et al. 1971: 70-73.

[40] Melo et al. 2016: 842.

[41]  Casanova 1995: 61; Casanova 2001: 33; Casanova 2003: 180-182.

[42] Casanova 2001: 35-36.

[43] The final decision was reached after discussion with the book specialist Professor Nicholas Pickwoad.

[44] Araújo et al. 2015: 136.

[45] Appelbaum 1989: 67. In accordance with the concept introduced by Appelbaum: “The notion of re-treatability is one that is often more helpful in evaluating treatments than the idea of reversibility itself.”



Conceição Casanova (1,2,3), Ana Lemos (3), Rita Araújo (1,2,3), Isamara Carvalho (1,4)

(1) Departamento de Conservação e Restauro, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa (Conservation and Restoration Department, Faculty of Sciences and Technology, NOVA University of Lisbon), 2829-516, Monte de Caparica, Portugal.

(2) LAQV-REQUIMTE, Department of Chemistry, Faculty of Sciences and Technology, NOVA University of Lisbon, 2829-516 Monte de Caparica, Portugal

(3) Instituto de Estudos Medievais, Faculdade de Ciências Sociais e Humanas, Universidade Nova de Lisboa (Institute of Medieval Studies, Faculty of Social and Human Sciences, NOVA University of Lisbon), Avenida de Berna 26-C,1069-061 Lisbon, Portugal

(4) Laboratório de Restauração, Coordenadoria de Preservação, Fundação Biblioteca Nacional (Conservation-Restoration Laboratory, Preservation Coordination, National Library of Brazil), Av. Rio Branco 219, 20040-008, Rio de Janeiro, Brazil.


The authors would like to thank the director and staff of the Library of Mafra National Palace, who are responsible for the precious Book of Hours collection and allowed us full access to study them.

We are also grateful to Professor Maria João Melo, president of the Conservation Restoration Department and responsible for the Scientific Lab, and the team member who supervised data interpretation and the analyses results. We thank the two DCR master students, Joana Bulcão and Ana Luís, that participated in the scientific equipment operations and conservation tasks.

This study was supported by the Portuguese Science Foundation through the CORES Ph.D. Doctoral Programme in the Conservation and Restoration of Cultural Heritage with grant SFRH/BD/52314/2013 and by the Associated Laboratory for Sustainable Chemistry-Clean Processes and Technologies-LAQV, which is financed by national funds from FCT/MEC (UID/QUI/50006/2016) and co-financed by the ERDF under the PT2020 Partnership Agreement (POCI-01-0145-FEDER-007265).

Helsinki • Artefacta • 3 April 2019